Rift Segmentation: structural mapping of syn-rift successions between the Kerpini-Tsivlos and Mamoussia-Pirgaki Faults, Greece

(1)

Faculty of Science and Technology

MASTER’S THESIS

Study program/Specialization:

Petroleum Geosciences Engineering

Spring semester, 2017

Open Author:

Eirik Oppedal

(Author’s signature)

Faculty supervisor: Chris Townsend

Title of thesis:

Rift Segmentation: structural mapping of syn‐rift successions between the Kerpini‐Tsivlos and Mamoussia‐Pirgaki Faults, Greece

Credits (ECTS): 30

Keywords:

Greece Corinth Rift Syn‐rift

Kerpini‐Tsivlos Fault Mamoussia‐Pirgaki Fault Half‐graben

Structural Geology Transfer Faults

Number of pages: ………

+ enclosure: ……….

Stavanger, 15.06.2017

(2)

Copyright by Eirik Oppedal

2017

(3)

Rift Segmentation: structural mapping of syn-rift successions between the Kerpini-Tsivlos and Mamoussia-Pirgaki Faults, Greece

By

Eirik Oppedal, BSc.

Master’s Thesis

Presented to the Faculty of Science and Technology The University of Stavanger

The University of Stavanger June 2017

(4)

Acknowledgements

I would like to express my sincere gratitude to my supervisor Chris Townsend for his continuous support and guidance. His knowledge of the Corinth Rift and experience in the field has been invaluable during the field trips. I would also like to thank my co-supervisor Alejandro Escalona for his feedback during the semester. Further thanks go to my colleagues, Asbjørn Veiteberg and Herman Birkeland, for their company and encouragement throughout the entire project.

(5)

Abstract

Rift Segmentation: structural mapping of syn-rift successions between the Kerpini-Tsivlos and Mamoussia-Pirgaki Faults, Greece

Eirik Oppedal

The University of Stavanger Supervisor: Chris Townsend

The Corinth Rift of Central Greece is currently active in the Gulf of Corinth. A series of rotated fault blocks from the early rift stages are preserved onshore in the southern part of the rift. These are well-exposed in incised river valleys, allowing detailed studies of normal faulting and associated syn-rift sedimentation. The study area for this project is limited between the Kerpini- Tsivlos and Mamoussia-Pirgaki Faults. This is an area where lateral correlation of faults and sedimentary packages is challenging, evident by a great variety in previous interpretations. Several major faults cannot be traced directly across the river valleys, and there is an ongoing debate about whether these faults terminate or are displaced laterally by relay or transfer structures. Through detailed structural mapping assisted by 3D modelling software, this study has investigated the along-strike continuity of faults and depocentres. Furthermore, it has investigated the relative age relationships between faulting and sedimentation in order to contribute to the understanding of the rift evolution.

Extensive NNE-SSW trending intervals of miscorrelation are identified in three different river valleys, indicating the presence of underlying transfer faults which are perpendicular to the graben- bounding faults. The transfer faults are segmenting the rift, and significant variations in bedding geometry/dip and fault throw between individual segments suggest that they have evolved individually. The structural evolution of the Corinth Rift is thus more complex than previously assumed. Fault activity in the early rift stages was broadly distributed across a number of faults, and the southern rift margin has generally migrated northwards through time.

(6)

Figure 1: Tectonic plate configuration in the Eastern Mediterranean Sea. Black arrows indicate the direction of plate movement. Modified after Wood (2013). ... 4 Figure 2: Structural overview of the Corinth Rift and its southern margin. The study area is marked with a red box. Modified after Wood (2013). ... 5 Figure 3: Structural map of the central northern Peloponnese Peninsula. The study area is

highlighted in brighter colours. Cross section B-B` is displayed in Figure 4. Modified after Ford et al. (2010). ... 7 Figure 4: Cross section B-B` from Figure 3. The study area is marked with a red box. Modified after Ford et al. (2010). ... 8 Figure 5: Wheeler diagram showing the stratigraphic correlations along the N-S profile in Figure 4.

The study area is marked in a red box. Modified after Ford et al. (2010). ... 8 Figure 6: Evolution of the Corinth Rift in four steps, showing a northward fault propagation and the proposed linkage to the Chelmos detachment fault. 1: Basement rock. 2: Syn-rift sediments.

3: Micro earthquakes from Rietbrock et al. (1996), recorded 15 km west of this cross section. Modified after Sorel (2000). ... 9 Figure 7: Conceptual model of transfer faults in a developing rift. Modified after Lister et al. (1986).

... 10 Figure 8: Conceptual model of a relay ramp. Modified after Athmer et al. (2010). ... 11 Figure 9: Structural map by Dahman (2015) with extensive N-S transfer faults. ... 12 Figure 10: Evolution of the southern margin of the Corinth Rift. Modified after Collier and Jones (2004). ... 14 Figure 11: Definition of roundness and sphericity used for field descriptions. Modified after

Krumbein and Sloss (1956). ... 17 Figure 12: Structural map of the study area. ... 20 Figure 13: Structural map of the study area, highlighting the 3 main segments into which it can be divided. ... 21 Figure 14: Generalized stratigraphic overview of the study area (roughly divided into 3 segments), illustrating the boundaries between the different units. The base of the Fluvial Conglomerates in the central segment is unexposed, lying in the sub-surface. ... 22 Figure 15: Structural map showing the location of the outcrops presented in Sub-chapter 3.2. ... 23 Figure 16: Pindos Basement. Outcrop location is marked on Figure 15. ... 24 Figure 17: Northern Basal Alluvial Conglomerates. The outcrop location is marked on Figure 15. 25 Figure 18: Southern Basal Alluvial Conglomerates. Outcrop location is marked on Figure 15. ... 27 Figure 19: Rose diagram showing paleo flow indicators for the Fluvial Conglomerates. ... 28 Figure 20: Conglomeratic channel bodies and floodplain deposits. Some (but not all) of the channels are highlighted. The outcrop location is marked on Figure 15. ... 29 Figure 21: Rose diagram showing paleo flow indicators for the Upper Alluvial Conglomerates. .... 30 Figure 22: Upper Alluvial Conglomerates at the peak of Mt. Petrouchi. The outcrop location is

marked on Figure 15. ... 31 Figure 23: Upper Alluvial Conglomerates unconformably overlying the Basal Alluvial

Conglomerates (the unconformity is not shown here). The location of the photo is marked in Figure 15. ... 32 Figure 24: Grey Breccia/conglomerate. a) Sharp contact between Basal Alluvial Conglomerates

(base) and Grey Breccia/conglomerates (top). b) Characteristic “sponge-like” matrix. The photo is from the same unit as in Figure 25, in a part with more angular clasts. The locations of the photos are found in Figure 15. ... 33

(9)

Figure 26: Grey Breccia/conglomerates situated within the Basal Alluvial Conglomerates. The locations of the photos are found in Figure 15. ... 35 Figure 27: Structural map showing the four different viewpoints from which Section 1 is photoed.

... 36 Figure 28: Section 1, View 1. A: original photo, B: observed lithological units and bedding, C:

structural interpretation. The location of the photoed section is found in Figure 27. ... 38 Figure 29: Section 1, View 2. A: original photo, B: observed lithological units and bedding, C:

structural interpretation. The location of the photoed section is found in Figure 27. ... 42 Figure 31: Conceptual model which can explain the folded Upper Alluvial Conglomerates in the

hangingwall of MSF3. The bedding dip increases towards the fault tip. ... 43 Figure 32: Section 1, View 4. A: Original photo, B: observed lithological units and bedding, C:

Structural interpretation. The location of the photoed section is found in Figure 27. ... 45 Figure 33: a) Prinos Fault. b) Poorly exposed Prinos unconformity. c) Toriza Fault. ... 46 Figure 34: Complete interpretation of Section 1 on a satellite image from Google Earth. The

horizontal and vertical scales are not proportional. The section is a compilation of the four views marked on Figure 27. ... 46 Figure 35: Structural map showing the three different viewpoints from which Section 2 is photoed.

... 47 Figure 36: Section 2, View 1. A: original photo, B: observed lithological units and bedding, C:

structural interpretation. The location of the photoed section is found in Figure 35. ... 53 Figure 39: Full interpretation of Section 2 on photo from Google Earth. The horizontal and vertical scales are not proportional. The section is a compilation of the three views marked on Figure 35. ... 54 Figure 40: Structural map showing the four different viewpoints from which Section 3 is photoed.

... 55 Figure 41: Section 3, View 1. A: original photo, B: lithology and structural interpretation. The

location of the photoed section is found in Figure 40. ... 56 Figure 42: Section 3, View 2. A: original photo, B: observed lithological units and bedding (plus the South Graben Fault in the distance), C: structural interpretation. The location of the photoed section is found in Figure 40. ... 58 Figure 43: Section 3, View 3. A: original photo, B: bedding, C: lithological and structural

interpretation. The location of the photoed section is found in Figure 40. ... 60 Figure 44: Section 3, View 4. A: original photo, B: observed lithological units and bedding (plus the South Graben Fault), C: structural interpretation. The location of the photoed section is found in Figure 40. ... 62 Figure 45: Full interpretation of Section 3 on photo from Google Earth. The horizontal and vertical scales are not proportional. The section is a compilation of the four views in this sub- chapter. ... 63 Figure 46: Structural map showing the three viewpoints from which Section 4 is photoed. ... 64 Figure 47: Section 4, View 1. A: bedding and lithological units, B: structural interpretation. The

location of the photoed section is found in Figure 46. ... 66 Figure 48: Section 4, View 2. A: original photo, B: observed lithological units and bedding, C:

structural interpretation. The location of the photoed section is found in Figure 46. ... 70

(10)

Figure 50: Outcrop of the northernmost north-dipping conglomerates in Section 4, which might be a late NE-prograding unit. The outcrop location is marked on Figure 49 ... 71 Figure 51: Google Earth photo with interpretation of Section 4. The horizontal and vertical scales are not proportional. The section is a compilation of the three views in the previous sub- chapter. ... 71 Figure 52: Structural map showing the viewpoint from which Section 5 is photoed. ... 72 Figure 53: Section 5. A: original photo, B: observed lithological units and bedding, C: structural

interpretation. The location of the photoed section is found in Figure 52. ... 74 Figure 54: Zoomed-in photo of the uppermost conglomerate beds in Section 5, interpreted as the

Upper Alluvial Conglomerates and differentiated from the Basal Alluvial Conglomerates.

... 75 Figure 55: Comparison of all five studied valley sections (approximately to scale) and correlation of faults (dashed black lines). The location of each section is marked on Figure 56. ... 77 Figure 56: Structural map showing the location of the five studied valley sections and Figure 57. .. 78 Figure 57: View down Krathis River. The eastern side is downthrown approximately 600 metres by a NNE-SSW striking fault. The location of the photo is marked on Figure 56. ... 79 Figure 58: Illustration of how the faults are extrapolated. Black fault line = observed fault. Red fault line = inferred fault. ... 81 Figure 59: Image from Petrel where faults are not correlatable across Ladopotamos River. The

location of the modelled faults is found in Figure 58. ... 82 Figure 60: Mamoussia-Pirgaki Fault trace (marked with arrows) in Google Earth. The location of the photo is found in Figure 58. ... 83 Figure 61: Comparison of the suspected North Graben Fault trace in Google Earth (a) with model fault trace in Petrel (b). The modelled plane dips 45°S. ... 84 Figure 62: Comparison of suspected South Graben Fault trace in Google Earth (a) with model fault trace in Petrel (b). The modelled planes dip 40°N. ... 84 Figure 63: a) Trace of the modelled eastern Kerpini-Tsivlos Fault, with plane dipping 42°NNW. b) Modelled central Kerpini-Tsivlos Fault and branched fault by the town of Souvardo. Black dots show the locations of observed fault contacts in the field. The locations of a) and b) are seen in Figure 58. ... 85 Figure 64: a) Suspected fault trace in Google Earth. b) Modelled fault plane in Petrel. The location of the figures are marked on Figure 58 ... 86 Figure 65: Modelled basement unconformity (dipping 35°SSW) in the easternmost Kerpini-Tsivlos Fault Block. The location of this figure can be seen in Figure 58... 87 Figure 66: Structural map showing the locations of the three interpreted N-S cross sections. ... 88 Figure 67: Western cross section (A-A’). The location of the cross section is marked on Figure 66. 89 Figure 68: Central cross section (B-B’). The location of the cross section is marked on Figure 66. . 90 Figure 69: Eastern cross section (C-C’). The location of the cross section is marked on Figure 66. . 90 Figure 70: Cumulative displacement plot of the western (Figure 67), central (Figure 68) and eastern (Figure 69) cross sections. The Kerpini-Tsivlos Fault is located by the zero-point for all three sections. ... 91 Figure 71: Structural map with proposed transfer faults (solid red lines) and other identified fault steps (dashed red lines). ... 94 Figure 72: Structural map with alternative interpretation involving straight transfer faults (red

lines). In this model the transfer faults themselves are stepping in the Ladopotamos River.

... 95 Figure 73: Proposed segmentation of the study area. The individual segments are separated by the inferred transfer faults (solid red lines). ... 96

(11)

Figure 75: E-W cross section (B-B’) showing the relationship between basement and sedimentary infill across proposed transfer faults. The location of the cross section is marked on Figure 73. ... 97 Figure 76: E-W cross section (C-C’) showing the relationship between basement and sedimentary

infill across proposed transfer faults. The location of the cross section is marked on Figure 73. ... 97 Figure 77: Structural map highlighting the area of the proposed evolution model, and the location of the three cross sections discussed in this sub-chapter. ... 100 Figure 78: Simplified structural maps illustrating the proposed four-stage evolution of the study area.

Solid red lines: active faults. Dashed red lines: active faults pre/post period of maximum displacement. Solid black lines: inactive faults. Dashed black lines: transfer faults. Dashed blue lines: rivers. ... 101 Figure 79: Generalized chronostratigraphic diagram showing the relative timing of faulting and

sedimentation, and the four stages of evolution. Active faults are shown as thick solid lines.

... 102 Figure 80: Proposed evolution of the western part of the study area. Solid red lines: active faults.

Dashed red lines: active faults pre/post period of maximum displacement. Black lines:

inactive faults. The location of the section is marked on Figure 77. ... 103 Figure 81: Proposed evolution of the central part of the study area. Solid red lines: active faults.

inactive faults. The location of the section is marked on Figure 77. ... 104 Figure 82: Proposed evolution of the eastern part of the study area. Solid red lines: active faults.

inactive faults. The location of the section is marked on Figure 77. ... 105

(12)

List of tables

Table 1: General characteristics of the northern Basal Alluvial Conglomerates... 25

Table 2: General characteristics of the southern Basal Alluvial Conglomerates. ... 26

Table 3: General characteristics of the Fluvial Conglomerates. ... 28

Table 4: General characteristics of the Upper Alluvial Conglomerates. ... 30

Table 5: General characteristics of the Grey Breccia/conglomerate. ... 33

(13)

Chapter 1 – Introduction

The interaction between the African and Anatolian tectonic plates initiated back-arc extension in the Oligocene (Jolivet et al., 1994; Le Pichon & Angelier, 1979), which resulted in the development of several active rifts. The Corinth Rift of Central Greece is one of them. It is among the world’s most active rifts, and initiated some 5 million years ago (Doutsos & Piper, 1990;

Leeder et al., 2008; Ori, 1989). A series of ESE-WNW striking rotated fault blocks from the early stages of this rift is preserved in the northern Peloponnese Peninsula, south of the Gulf of Corinth.

Incisive north-trending river valleys provide exposures of these fault blocks, as they cut near perpendicular to strike of the main structures. This allows for detailed studies of normal faulting and associated syn-rift sedimentation, and it provides a unique opportunity to understand the development of ancient rift systems. Furthermore, this area can be used as an extensional basin analogue for hydrocarbon exploration (e.g. on the Norwegian Continental Shelf). The northern Peloponnese has thus been studied and mapped in various detail (Collier & Jones, 2004; Dahman, 2015; Ford et al., 2013; Hadland, 2016; Rohais et al., 2007; Sigmundstad, 2016; Skourtsos &

Kranis, 2009; Stuvland, 2015; Syahrul, 2014; Wood, 2013), and as a result, several different interpretations of the evolution and current structural-stratigraphic configuration of the rift system have developed.

The study area for this project is limited between the Vouraikos and Krathis River Valleys.

Approximate southern and northern boundaries are the Kerpini-Tsivlos and Mamoussia-Pirgaki Faults respectively. This is an area where outcrops and elevated viewing points are not easily accessible, and it is yet to be mapped in detail. Lateral correlation of faults and sedimentary packages is challenging, evident by a great variety in previous interpretations. Several major faults cannot be traced directly across the river valleys, and there is no wide agreement on whether individual faults simply terminate in the valleys, or are linked to parallel faults by relay or transfer structures. Even though most recent publications acknowledge the presence of fault-linking structures in the Corinth Rift, there has been little emphasis on understanding how abundant and extensive the structures are (some exceptions: Dahman (2015); Ford et al. (2016); Ford et al.

(2013); Wood (2013)).

(14)

The purpose of this study is firstly to investigate the E-W continuation of faults and depocentres in the study area through detailed structural mapping combined with 3D modelling software.

Secondly, it aims to present evidence for the relative age relationships between faulting and sedimentation in order to contribute to the understanding of the rift evolution.

(15)

1.1 Geological framework 1.1.1 Regional geology

The tectonic setting in the Eastern Mediterranean Sea is influenced by the interaction of the African, Anatolian, Eurasian and Arabian Plates (Figure 1). The Gulf of Corinth is located in the north-western part of the Anatolian Plate. This plate is bounded by the Hellenic Trench in the southwest (where the African Plate is subducting below the Anatolian Plate), by the right lateral North Anatolian Fault in the north (which separates it from the Eurasian Plate), and by the left lateral East Anatolian Fault in the east (which separates it from the Arabian Plate). The Anatolian Plate can be subdivided into the smaller Aegean and Anatolian Plates (Jackson, 1994), but the boundary between them is not entirely agreed upon. Papazachos (1999) defines it in Western Turkey, while (Scott, 1981) infers it below the Mediterranean Sea.

The tectonic evolution of the Eastern Mediterranean Sea is dominated by two main factors:

1. North-eastward subduction of the African Plate at the Hellenic Trench

2. Northward continental collision of the Arabian Plate into the Anatolian Plate in Eastern Turkey (Taymaz et al., 2007)

The continental collision forces an anti-clockwise rotation of the Anatolian Plate along the East and North Anatolian Faults, pushing it west-southwest towards the Hellenic Trench. Back-arc extension in the Aegean Sea and Southern Greece is a result of slab pull related to the Hellenic subduction, and the anti-clockwise rotation of the Anatolian Plate. The back-arc extension initiated in the Oligocene (Gautier et al., 1999; Jolivet et al., 1994; Le Pichon & Angelier, 1979), while the rotation of the Anatolian Plate initiated in the Pliocene (Armijo et al., 1996).

Within the Anatolian extensional system, the Corinth Rift is the most active among a number of WNW-ESE trending rifts. Based mainly on micropaleontological dating, its initiation is estimated to be at around 5 Ma (Doutsos & Piper, 1990; Leeder et al., 2008; Ori, 1989). The Corinth Rift was superimposed on Hellenide thrust sheets which where emplaced westwards during the Cretaceous to Miocene (Richter, 1976).

(16)

Figure 1: Tectonic plate configuration in the Eastern Mediterranean Sea. Black arrows indicate the direction of plate movement. Modified after Wood (2013).

(17)

1.1.2 Structural and stratigraphic overview

The Gulf of Corinth is a 115 kilometre WNW-ESE oriented elongated graben, separating the Peloponnese Peninsula in the south from mainland Greece in the north (Figure 2). Its northern margin is dominated by south-dipping faults, and its southern margin by north-dipping faults. The rift system is currently active in the Gulf of Corinth, while inactive faults from the early rift stages are preserved onshore in the northern Peloponnese Peninsula (Figure 3). The early rift is characterized by a series of rotated fault blocks with associated (syn-rift) sedimentary infill in half- grabens (Figure 4). Incised river valleys (trending SSW-NNE) provide excellent exposures of these fault blocks, as they cut perpendicular to the strike of the main structures. This allows for detailed studies of the entire southern rift margin, from the town of Kalavryta in the south to the Gulf of Corinth in the north. All the major graben bounding faults are north-dipping, with dip angles in the range of 40-60°.

Figure 2: Structural overview of the Corinth Rift and its southern margin. The study area is marked with a red box.

Modified after Wood (2013).

In the Kalavryta area, the so-called Pindos thrust sheet constitutes the dominant pre-rift stratum. It is mostly composed of Upper Triassic-Jurassic, highly deformed, metamorphosed carbonates (Skourlis & Doutsos, 2003), and is one of several Hellenide thrust sheets which were emplaced westwards during the Cretaceous to Miocene. The younger syn-rift infill is described in various detail by several researchers. A publication by Ford et al. (2013) describes the southern rift margin

(18)

as a whole, with the aim to understand the tectono-sedimentary evolution of the Western Corinth Rift. They have classified three main stratigraphic groups (Figure 5) within the syn-rift deposits:

Lower Group: The Lower Group is dominant in the southern part of the rift, extending from the Kalavryta Fault Block in the south to the Mamoussia-Pirgaki Fault Block in the north. It consists generally of (terrestrial) coarse conglomeratic fluvial/alluvial to fine-grained lacustrine successions.

Middle Group: While the Lower Group is characterized by terrestrial deposits, the Middle Group consists of alluvial and Gilbert-type fan deltas that have prograded northwards into a brackish/marine environment. Laterally equivalent distal turbidites and hemipelagic suspension deposits can be found alongside the prograding deltas. The Middle Group is separated from the Lower Group by an erosional unconformity, and it is dominantly confined to the Mamoussia- Pirgaki Fault Block. Smaller portions of the hemipelagic and turbiditic deposits extend northwards to the Helike Fault Block.

Upper Group: The Upper Group is mainly deposited offshore, and it includes the sediments currently being deposited in the active part of the rift. Onshore it is characterized by conglomeratic Gilbert-type deltas, concentrated along the hangingwall of the Helike Fault where they unconformably overlie the Middle and Upper groups and build into the Gulf of Corinth.

(19)

Figure 3: Structural map of the central northern Peloponnese Peninsula. The study area is highlighted in brighter colours. Cross section B-B` is displayed in Figure 4. Modified after Ford et al. (2010).

(20)

Figure 4: Cross section B-B` from Figure 3. The study area is marked with a red box. Modified after Ford et al.

(2010).

Figure 5: Wheeler diagram showing the stratigraphic correlations along the N-S profile in Figure 4. The study area is marked in a red box. Modified after Ford et al. (2010).

(21)

1.2 Previous work

The geometry of the Corinth Rift has been subject to debate for decades. Doutsos and Piper (1990) proposed that the normal faults in the Peloponnese Peninsula are of a listric nature, but the lack of evidence supporting such a model have seen most researchers favour a planar fault model (e.g.

Ford et al. (2013); Moretti et al. (2003); Rohais et al. (2007); Westaway (2002)). Doutsos and Poulimenos (1992) suggested that the normal faults at the surface link to an underlying major low- angle normal fault. Later, Flotté and Sorel (2001) and Chery (2001) developed on that idea, and proposed that the rift is underlain by a north-dipping crustal detachment fault which is exposed in the southernmost part of the rift system (Chelmos Fault, Figure 6). Studies of focal mechanisms (Rietbrock et al., 1996) from the Aigion earthquake in 1995 have also been used to support a model involving an active low-angle crustal detachment, as the cluster of recorded micro earthquakes show a north-dipping zone of seismicity below the gulf. A dominance of north-dipping faults and the suggested detachment has raised the question as to which the rift is symmetrical or asymmetrical (e.g. Jolivet et al. (2010); McNeill et al. (2005); Moretti et al. (2003))

Figure 6: Evolution of the Corinth Rift in four steps, showing a northward fault propagation and the proposed linkage to the Chelmos detachment fault. 1: Basement rock. 2: Syn-rift sediments. 3: Micro earthquakes from Rietbrock et al. (1996), recorded 15 km west of this cross section. Modified after Sorel (2000).

(22)

The Corinth Rift system is segmented, evident by large (up to kilometre-scale) steps of major faults along approximately N-S trends. Ghisetti and Vezzani (2005) suggest that the rift segmentation is controlled by pre-existing structures in the underlying pre-rift Pindos Basement. Steps are especially prominent in river valleys, where several faults cannot be directly correlated across.

There is so far no wide agreement on whether individual faults simply terminate in the valleys, or are linked to parallel faults by relay or transfer structures (conceptual models: Figure 7 and Figure 8). Both structures are identified in other rift systems, e.g. the Rio Grande Rift (Mack & Seager, 1995), the Reconcavo Graben (Milani & Davison, 1988), the Suez Rift (Moustafa, 1996) and the East African Rift (Morley et al., 1990).

Ford et al. (2013) propose that each fault step is caused by an individual cross fault (e.g. in their Kerpini and Mamoussia Faults, Figure 3) or relay zone. The latter is supported by Wood (2013).

Dahman (2015) identified an extensive N-S interval of miscorrelation in Vouraikos River, and prefers a model involving several kilometres long transfer faults (Figure 9).

(23)

Figure 8: Conceptual model of a relay ramp. Modified after Athmer et al. (2010).

(24)

(25)

A general consensus has been reached on the idea of northward migration of fault activity (Ori, 1989; Sorel, 2000; Collier and Jones, 2003; Flotté et al., 2005; Ford et al., 2013), from presently inactive faults south of the town of Kalavryta to the active rift in the Gulf of Corinth. This implies that the syn-rift sediments are progressively younger towards the Gulf of Corinth. Although the general northward migration is agreed upon, the relative timing between individual faults remains unclear.

Sorel (2000) suggests a sequential northward fault migration, where the displacement of an abandoned fault is always transferred to a fault further north (Figure 6). This is disputed by other authors (Collier & Jones, 2004; Rohais et al., 2007) who propose a model where fault activity is distributed across several active faults (Figure 10), where the “zone” of active faults propagate northwards. This model involves simultaneous syn-rift deposition in different half-grabens. Based on U/Th dating, Causse et al. (2004) suggest that the Dhoumena Fault was active at around 0.125 Ma. This is significantly out of sequence with the model by Sorel (2000), and potentially supports the model by Collier & Jones (2004) and Rohais et al. (2007).

(26)

Figure 10: Evolution of the southern margin of the Corinth Rift. Modified after Collier and Jones (2004).

(27)

1.3 Thesis motivation and objectives

The Corinth rift serves as an important analogue for structural uncertainty in hydrocarbon exploration (in extensional basins, e.g. on the Norwegian Continental Shelf). The geometries of faults and the linkages between them control the distribution of hanging wall reservoirs. The way a fault is extrapolated along strike heavily influences reservoir volumes and the distribution of facies. Fault discontinuities (i.e. steps), as seen in the northern Peloponnese Peninsula, are not easily identifiable on 2D seismic during exploration. When not identified, the volumes of a prospect can be overestimated as displacement minima related to relay and transfer structures often define structural spill points. Being able to predict/identify along-strike discrepancies would thus have significant implications for the economic viability of a prospect. Furthermore, insight in the evolution of rift systems is critical for understanding the timing of petroleum system elements in an exploration setting.

The purpose of this study is firstly to investigate the E-W continuation of faults and depocenters in the study area through detailed structural mapping. Secondly, it aims to present evidence for the relative age relationships between faulting and sedimentation in order to contribute to the understanding of the rift evolution. All the objectives for the thesis are defined as follows:

 Make a detailed interpretation of 5 valley sections, and investigate the correlation between them.

 Determine the direction of sediment transport (paleo flow) within different lithologies, and consider their source area and down-stream rock equivalents.

 Compile structural and stratigraphic evidence in order to establish relative age relationships.

 Construct a 3D structural-stratigraphic model describing the present day configuration.

 Propose a structural-sedimentary evolution model of the study area.

(28)

Chapter 2 – Methodology

The methodology for this project can be divided into three parts; pre-fieldwork, fieldwork and post-fieldwork.

2.1 Pre-fieldwork

In order for the field trips to be as rewarding as possible, and for the objectives to be met, thorough preparations were made. Published literature was review in order to identify locations of particular interest, and to get a general understanding of the different existing interpretations. A preliminary structural map based on the reviewed literature was made prior to the first field trip in July/August, providing an overview of what should be investigated. Hiking and driving routes for each day were also prepared, ensuring no time was wasted in the field. Prior to the second field trip (April, 2017) an evaluation of the data from the first trip was done, uncovering lacking data and new areas of interest.

2.2 Fieldwork

The fieldwork was carried out over a total of 25 days, and consisted of collecting data in the form of:

• Sedimentological descriptions of outcrops.

• Photos and sketches of valley exposures (including use of a DJI Phantom 3 Drone equipped with camera).

• Locations of faults and lithological contacts.

• Strike and dip measurements of faults, bedding planes and unconformities.

• Paleo current measurements from imbrication, cross bedding and channel geometries.

The main focus during the first field trip was to describe the valley exposures as detailed as possible, in order to have a solid initial framework of the study area. The second field trip consisted of mapping specific areas between the valleys in greater detail, to test the theories developed in the months after the first trip.

(29)

preferred clast orientation. The clast size is estimated by measuring the long axis of the 10 largest clasts within a square metre which is representative for the unit being described. The average measurement of 5-10 square metres is the clast size presented in this thesis. The clast sphericity and roundness are described according to the definition in Figure 11.

Figure 11: Definition of roundness and sphericity used for field descriptions. Modified after Krumbein and Sloss (1956).

2.3 Post-fieldwork

The first step after both field trips was to organize the collected data, making them easily accessible for interpretation. GPS points were imported to a geographic information system (GIS) database, in order to geo-reference all the collected data on a map. The numerous photos of the valley exposures were merged into larger composite photos and interpreted in CorelDraw.

Interpretation of fault continuation between valleys (wherever challenging terrain prevented tracing in the field) were done by combining Google Earth and 3D structural modelling in Petrel.

The strike of faults observed in the field is often unknown, but some faults are clearly visible on

(30)

satellite images. Other structures are less visible, but several of the faults observed in the field tend to be located on boundaries between sparsely and densely vegetated areas, or simply on a small groove in the topography. During modelling in Petrel, it was tested how planar fault surfaces intersect a digital elevation model (DEM), and how those intersections compared to the features observed in Google Earth.

(31)

Chapter 3 – Field observations and interpretations

3.1 Introduction

The structure of the study area (Figure 12) is characterized by a majority of north-dipping faults, exposed in the Vouraikos, Ladopotamos, Potamia and Krathis River Valleys. Six major faults are here identified; the Kalavryta, Kerpini-Tsivlos, South Graben (not a previously used name), Dhoumena, Valimi and Mamoussia-Pirgaki Faults. The South Graben Fault might be considered as the eastward continuation of the Dhoumena Fault after a right-step in the Vouraikos River, but it is here interpreted as an individual fault. Although the Kalavryta and Dhoumena Faults with respective hangingwall lithologies are added to the map, they have not been studied in detail in this thesis.

The pre-rift metamorphosed carbonate (Pindos Basement) is well exposed in the eastern and north- western part of the study area, but thick successions of syn-rift sediments make up most of the exposed lithologies. The syn-rift sediments here correspond to the Lower Group by Ford et al.

(2013) which is characterized by terrestrial alluvial and fluvial conglomerates. The study area can roughly be divided into three different segments from south to north (Figure 13 and Figure 14), based on their exposed lithologies:

1. A southern segment dominated by coarse alluvial conglomerates. Two different alluvial units are identified (Basal and Upper Alluvial Conglomerates), separated by an unconformity. The Basal Alluvial Conglomerates are overlying the Pindos Basement.

There are also local exposures of a minor unit (Grey Breccia/conglomerate) which stands out from all the other lithologies because of its great amount of angular clasts.

2. A central segment characterized by a thick (>600 m) fluvial succession (Fluvial Conglomerates) with a buried base. It is conformable overlain by a massive unit of Upper Alluvial Conglomerates. The Pindos Basement is not exposed here.

3. A northern segment with well-exposed Pindos Basement. It is onlapped by an 80 metre thick unit of Basal Alluvial Conglomerates, followed by 300 metres of Fluvial Conglomerates.

The sediments in Segment 1 dip southwards towards the Kerpini-Tsivlos Fault, which is characteristic for the majority of fault blocks in the southern margin of the Corinth Rift. In

(32)

Segments 2 and 3 anomalous northward dips are observed, particularly to the west of Ladopotamos River.

The identified lithologies and structural features are described in more detail throughout this chapter. The structural features in five valley sections will be presented, captured by photos (both original and interpreted ones) from a number of different viewpoints. The photos are marked with key letters, where each letter has an associated description in bullet point format. The colour legend from Figure 12 is consistently used for the lithologies throughout this thesis.

Figure 12: Structural map of the study area.

(33)

Figure 13: Structural map of the study area, highlighting the 3 main segments into which it can be divided.

(34)

3.2 Lithologies

During the field work four lithological units were identified within the syn-rift half graben infill which unconformably overlies the Pindos Basement. They are all terrestrial clastics (dominantly conglomerates) and are differentiated based on brief sedimentological descriptions such as bed thickness, clast size, roundness, sphericity, sorting, grading and preferred clast orientation.

This sub-chapter will present general descriptions of all the observed units, including their stratigraphic relationship with the other units (see Figure 14 for stratigraphic overview) and outcrop photos (outcrop locations are marked on Figure 15).

Figure 14: Generalized stratigraphic overview of the study area (roughly divided into 3 segments), illustrating the boundaries between the different units. The base of the Fluvial Conglomerates in the central segment is unexposed, lying in the sub-surface.

(35)

Figure 15: Structural map showing the location of the outcrops presented in Sub-chapter 3.2.

3.2.1 Pindos Basement

The pre-rift basement unit (Figure 16) consists mainly of metamorphosed carbonates, with some local occurrences of shale. It is highly deformed due to the compressional regime under which the Hellenide thrust sheet was emplaced, and there are thus few reliable bedding surfaces for dip measurements. Its down-section extent is estimated by extrapolating the dip of unconformity surfaces separating it from the half graben sedimentary infill. This method relies on the assumption that the unconformity planes are approximately planar, although they may in reality have significant paleo topography.

(36)

Figure 16: Pindos Basement. Outcrop location is marked on Figure 15.

3.2.2 Basal Alluvial Conglomerates

Alluvial conglomerates are observed as the lowermost sedimentary infill at two separate locations;

in the north-western part of the study area (in the Vouraikos River) and in the hangingwall of the Kerpini-Tsivlos Fault. Stratigraphically, the two successions are not overlain by the same unit, and the southern succession is ~700 metres thicker than the northern one. Thus, it is unclear whether they can be considered as the same unit or not. The southern conglomerates are slightly coarser and poorer sorted, but in general their sedimentological character is similar (described below, Table 1 and 2). In this thesis they are interpreted as the same unit.

Northern Basal Alluvial Conglomerates

(37)

is cut by low-displacement (< 50 m) faults which is sealed by (or unexposed in) overlying strata.

It is overlain by the Fluvial Conglomerates, described in Sub-chapter 3.2.3. The southern boundary of this unit is abrupt, as it terminates laterally towards a 600 metre thick succession of Fluvial Conglomerates. The nature of this boundary will be discussed in Chapter 3.3. The general lithological description of this stratigraphic unit is found in Table 1

Table 1: General characteristics of the northern Basal Alluvial Conglomerates.

Characteristic Description

Clast size 13 cm

Sorting Moderate --> poor

Grading None

Bed thickness 10-30 m

Clast roundness Sub-rounded

Preferred clast orientation None

Clast sphericity Low --> moderate

Figure 17: Northern Basal Alluvial Conglomerates. The outcrop location is marked on Figure 15.

(38)

Southern Basal Alluvial Conglomerates

In the footwall of the South Graben Fault is an 800 metre succession of alluvial conglomerates overlying the Pindos Basement (Figure 18). Unlike the northern Basal Alluvial Conglomerates, its angular relationship with the basement is ambiguous. It appears to onlap the basement in the Vouraikos River, but the unconformity there is poorly exposed and the relationship can thus not be determined with certainty. To the very east in the study area, in the Krathis River, the beds are parallel to the unconformity. Locally, the Basal Alluvial Conglomerates are unconformably overlain by the Upper Alluvial Conglomerates.

The unit consistently dips 20-30°S towards the Kerpini-Tsivlos Fault, which conforms to the typical south-tilting observed in other fault blocks in the southern rift margin. Its bed thicknesses decrease up-section, suggesting that the unit is retrograding. The beds near its base are similar in thickness to those of the northern Basal Alluvial Conglomerates.

Although not supported by adequate sedimentological observations, the northern Basal Alluvial Conglomerates are here correlated to the lower beds of the southern Basal Alluvial Conglomerates.

They are considered as the first and most extensive deposits in a south-retrograding alluvial fan system. The lithological description of the southern Basal Alluvial Conglomerates is found in Table 2.

Table 2: General characteristics of the southern Basal Alluvial Conglomerates.

Clast size 15 cm

Sorting Poor Grading None

Bed thickness 5-30 m (decreasing up-section)

Clast sphericity Low --> moderate

(39)

Figure 18: Southern Basal Alluvial Conglomerates. Outcrop location is marked on Figure 15.

3.2.3 Fluvial Conglomerates

In the hangingwall of the South Graben Fault is a thick fluvial succession. The observable part of the succession is 600 metres thick, but as its base is not exposed the thickness of this unit is even greater. It is overlain by the Upper Alluvial Conglomerates (described in Sub-chapter 3.2.4). There

(40)

are no observable discordance between the two units, and the contact appears to be conformable.

In the north-western part of the study area the lower part of the Fluvial Conglomerates terminates laterally against the Pindos Basement and Basal Alluvial Conglomerates (Figure 20). The upper 300 metres of the succession extend further northwards and overlies the Basal Alluvial Conglomerates. The angular relationship between these two units is not determined with certainty, as there are few well-exposed beds within the Fluvial Conglomerates directly above the contact.

It is comprised of 1-5 metre thick bodies of well sorted matrix-supported conglomerates, and thinner beds of light brown/red silt and fine sand. Channel geometries (typically of a 15-25 m width) are commonly present within the conglomerates, and clast imbrications are often observable in outcrops. The finer sediments are interpreted as sequential floodplain deposits. The clast size variety between individual channels is believed to relate to different levels of flow energy within a braided river system. The succession is fining northwards, both in terms of clast sizes and bed thicknesses. The general lithological description can be seen in Table 3. Clast imbrications and the azimuth of exposed channel bodies indicate a dominant paleo flow towards northeast (Figure 19).

Table 3: General characteristics of the Fluvial Conglomerates.

Characteristic Description Clast size 5-10 cm (decreasing

northwards)

Sorting Very well (best sorted conglomerates)

Grading Usually none. Some

occasions of normal grading Bed thickness 1-5 metres

Clast roundness Sub-rounded --> rounded Preferred clast orientation Occasionally imbricated Clast sphericity Moderate --> high

Figure 19: Rose diagram showing paleo flow indicators for the Fluvial

Conglomerates.

(41)

Figure 20: Conglomeratic channel bodies and floodplain deposits. Some (but not all) of the channels are highlighted. The outcrop location is marked on Figure 15.

(42)

3.2.4 Upper Alluvial Conglomerates

The uppermost unit is a sharp-based 300 metre thick succession of very massive clast-supported conglomerates with bed thicknesses reaching 30-40 metres. Beds are separated by thin layers of siltstone or sandstone. It is more poorly sorted than the Fluvial Conglomerates, and contains few internal structures (there are some rare occasions of imbrication (Figure 21). It is seen unconformably overlying the Basal Alluvial Conglomerates in the south (Figure 22), and conformably overlying the Fluvial Conglomerates in the central part of the study area (there are no observable discordance between the two) (Figure 23). The unit is interpreted as a significant progradational alluvial fan sourced from the south. Its northern boundary is abrupt, and it is limited either by erosion or a south-dipping fault. The general lithological description of the Upper Alluvial Conglomerates can be seen in Table 4.

Table 4: General characteristics of the Upper Alluvial Conglomerates.

Clast size 10-15 cm

Sorting Moderate --> poor

Grading Usually none. Inverse grading observed locally.

Bed thickness 10-40 metres. Most massive unit in terms of bed thickness.

Preferred clast orientation None. Some rare cases of imbrication.

Clast sphericity Moderate Figure 21: Rose diagram showing paleo flow indicators for the Upper Alluvial Conglomerates.

(43)

Figure 22: Upper Alluvial Conglomerates at the peak of Mt. Petrouchi. The outcrop location is marked on Figure 15.

30 m

10 m 20 cm

N

S

(44)

Figure 23: Upper Alluvial Conglomerates unconformably overlying the Basal Alluvial Conglomerates (the unconformity is not shown here). The location of the photo is marked in Figure 15.

3.2.5 Grey Breccia/conglomerates (minor unit)

A distinct 10-30 metre thick grey unit of hard, consolidated, matrix-supported breccia to conglomerate is locally exposed in 4 valley sections. It has a sharp top and base (Figure 24a), and a characteristic “sponge-like” surface as pebbles have fallen out of its matrix (Figure 24b). The

(45)

also seen within the Basal Alluvial Conglomerates (Figure 26). They share the same characteristic matrix, grey colour and variety in clast roundness. The distance between each outcrop is large, and it is unclear whether they can be considered as one single unit or not. However, in this thesis it is presented as such. The general lithological description of this stratigraphic unit can be seen in Table 5.

Table 5: General characteristics of the Grey Breccia/conglomerate.

Clast size 10 cm

Sorting Very poor. Clasts range from pebble to boulder

size. Chaotic.

Grading None

Bed thickness 1-3 m

Clast roundness Sub-angular. Wide roundness range from angular to rounded.

Clast sphericity Elongated --> low

Figure 24: Grey Breccia/conglomerate. a) Sharp contact between Basal Alluvial Conglomerates (base) and Grey Breccia/conglomerates (top). b) Characteristic “sponge-like” matrix. The photo is from the same unit as in Figure 25, in a part with more angular clasts. The locations of the photos are found in Figure 15.

(46)

Figure 25: Grey Breccia/conglomerate on top of Basal Alluvial Conglomerates. The outcrop location is marked on Figure 15.

(47)

Figure 26: Grey Breccia/conglomerates situated within the Basal Alluvial Conglomerates. The locations of the photos are found in Figure 15.

(48)

3.3 Valley section 1: Vouraikos East

The Vouraikos River Valley is among the most studied sections in the Kalavryta area, as several lithologies, including the basement, are very well exposed. This sub-chapter offers photos, both original and interpreted ones, from four different viewpoints in order to capture most of the features in the valley side. All the viewpoints are marked on Figure 27, and the photos are here presented from north to south.

Figure 27: Structural map showing the four different viewpoints from which Section 1 is photoed.

(49)

3.3.1 View 1

View 1 (Figure 28) shows the very north-western part of the study area, where a series of south- tilted (domino) basement blocks crop out at the valley bottom. They are overlain by 80-100 metres of Basal Alluvial Conglomerates, followed by a 300 metre succession of Fluvial Conglomerates.

To the very south the basement and Basal Alluvial Conglomerates are not exposed, and 600 metres of Fluvial Conglomerates are overlain by a 200 metre package of massive Upper Alluvial Conglomerates.

Observations + interpretation:

a. The Basal Alluvial Conglomerates onlap the basement unconformity. It is here evident that the basement blocks experienced initial rotation prior to deposition of the earliest syn-rift sediments.

b. The Basal Alluvial Conglomerates are cut by low-displacement (<50 m) faults which are either sealed by (or unexposed in) the overlying strata.

c. The basement and Basal Alluvial Conglomerates terminate laterally towards Fluvial Conglomerates in the south. The contact between them dips to the south. Two possible scenarios could explain this:

1. The southern side is down-thrown (minimum 300 m) by a south-dipping fault, and Basal Alluvial Conglomerates are buried below the thick succession of Fluvial Conglomerates.

2. The basement and Basal Alluvial Conglomerates are eroded by incised rivers, and the sharp lateral change is an unconformity.

(50)

Figure 28: Section 1, View 1. A: original photo, B: observed lithological units and bedding, C: structural interpretation. The location of the photoed section is found in Figure 27.

(51)

3.3.2 View 2

From this view (Figure 29) one can observe the Basal Alluvial Conglomerates from a different perspective, and how they terminate towards the thick succession of Fluvial Conglomerates with a buried base. 200 metres of Upper Alluvial Conglomerates are overlying the Fluvial Conglomerates, and they are here clearly differentiated. The contact between them is stepping down to the south.

a. The northern boundary of the Upper Alluvial Conglomerates is here controlled by erosion, as the unit extrapolates into mid-air.

b. The Upper Alluvial Conglomerates are offset by 2 clear south-dipping faults, previously named Mega Spilio Fault 3 (MSF3) and Mega Spilio Fault 2 (MSF2) by Dahman (2015).

Their displacement is 250 metres and 100 metres respectively. MSF2 terminates towards MSF3. A possible third south-dipping fault slightly offsets the uppermost beds.

c. The sharp contact between the Fluvial and Basal Alluvial Conglomerates appears to dip to the north. However, this is implausible as the sediments on the southern side of the contact are younger. The presence of either a south-dipping fault or unconformity both seem likely.

(52)

(53)

3.3.3 View 3

This view (Figure 30) offers a better perspective of the faulted Upper Alluvial Conglomerates seen in View 2, and it is highly oblique to the south-dipping Mega Spilio faults. It also captures the southern Basal Alluvial Conglomerates, although covered in vegetation.

a. The Upper Alluvial Conglomerates in the hangingwall of MSF3 are slightly folded. As the footwall sediments of MSF3 (at the top of Mt. Psili Rachi) are apparently sub- horizontal, the folded beds in the hangingwall might be a result of increasing fault slip along MSF3 towards the northwest (conceptual model, Figure 31). This fold geometry has previously been interpreted as a hangingwall syncline caused by fault-propagation folding or normal drag along a north-dipping fault (Dhoumena fault, Ford et al. (2013) and Wood (2013)).

b. There is evidence of a south-dipping fault (Mega Spilio Fault 1 (MSF1), Dahman (2015)), as a thick package of Upper Alluvial Conglomerates is downthrown 100 metres directly south of the Mega Spilio Monastery (marked on Figure 30). The thickness of this package is similar to the package at the top of Mt. Psili Rachi. The monastery is thus believed to sit on the fault plane of MSF1.

c. South of MSF1 is a 600 metre wide zone of dense vegetation and poor exposure. The contact between the Fluvial and Upper Alluvial Conglomerates is here not identified.

However, Fluvial Conglomerates are observed at the base, and Upper Alluvial

Conglomerates are observed near the mountain top, thus the contact has to be somewhere in between. The outcrops in this zone are generally dipping 10-20°N.

d. South of the vegetated zone (c) is an 800 metre succession of Basal Alluvial

Conglomerates with beds consistently dipping 25-30°S all the way to the Kerpini-Tsivlos Fault. No beds are traceable across a small incision in the valley side, and it is interpreted to represent the trace of the major north-dipping South Graben Fault. The displacement of this fault is uncertain, but it is estimated to be in region of 600-1500 metres.

e. On the mountain top, 200 metres south of the inferred South Graben Fault, is a small outcrop of Grey Breccia/conglomerate. It terminates southwards towards a poorly exposed north-dipping fault. A better exposure of this fault will be shown in Section 2.

(54)

(55)

Figure 31: Conceptual model which can explain the folded Upper Alluvial Conglomerates in the hangingwall of MSF3. The bedding dip increases towards the fault tip.

(56)

3.3.4 View 4

View 4 (Figure 32), the southernmost studied part of Vouraikos River, exhibits a very thick succession of Basal Alluvial Conglomerates dipping 25-30°S towards the Kerpini-Tsivlos Fault.

Its northern contact with the younger Fluvial and Upper Alluvial Conglomerates (the inferred South Graben Fault) is also shown here. At the base of the succession is a small rotated basement inlier (hidden behind the foreground in this photo). This inlier has previously been referred to as the Prinos Inlier (Ford et al., 2013; Dahman, 2015).

a. The Prinos Inlier stretches for 300 metres and can be traced, from a fault contact (Figure 33a) in the north to an unconformity contact (Figure 33b) in the south, along the road in the bottom of the valley. The measured dip of the fault and unconformity is 20° and 48°

respectively, indicating that the inlier is south-tilted. The unconformity exposure is rather poor, and its measured dip might not be reliable. If it is indeed dipping more than 40°, it would imply that the Basal Alluvial Conglomerates onlap the basement, as seen further north in View 1.

b. 150 metres south of the Prinos Inlier is another 30 metre small basement outcrop (Figure 33c), interpreted as being part of the footwall of the Toriza Fault identified by Dahman (2015).

c. The massive beds of Mt. Toriza abruptly terminate northwards against a north-dipping planar slope in the valley side, interpreted as a fault and its associated fault plane.

d. The beds of Mt. Toriza has previously been used as a prime example of syn-sedimentary tilting and progressive up-section dip decrease. However, the dip change observed here is very abrupt (from 27° to 10°), suggesting the presence of an unconformity separating two phases of alluvial conglomerate deposition. The unconformity is only visible in the upper 100 metres of the succession before being covered in vegetation, and so the thickness of the second-phase alluvial unit is uncertain. The overlying unit is here interpreted as the Upper Alluvial Conglomerates, which are overlying the Fluvial Conglomerates further north.